Information Technology Reference
In-Depth Information
uncertainties making them possible [2] and the consequential mistrust in the
implementation of local drought management acts [3]. Emergence of controversies is
also fostered by the diversity of ways participants to drought committee meetings can
assess drought situation by themselves: different places (e.g. where their well is
located, or the bridge where they cross the river is) and different resources
(groundwater or surface water). Each user comes also with his/her own indicator to
characterize drought: water level but also length of riverbed without water.
All come then with their own view gathered in the single possible assessment:
drought/No drought. Drought committee will rather ends up with a negotiated
assessment. Implementation and respect of rules which is formally generated by this
assessment will then depends on the adhesion of water users to the assessment, even
more with the weakness of means for control. This raises an issue of effectiveness and
fairness of these drought management acts according to the diversity of possible
scenarios for implementing them. These initial empirical studies could make clear this
concern of controversies in implementation while it is supposed to have all the rigor
of science. To go further, we needed some simulation tool to explore contrasted
scenarios of implementation of local drought management policy.
To create this tool, we took a pragmatic stance: explaining dynamics generated by
this policy with “situated action” instead of planned action [4-5]. Agent Based
Modelling has a suitable format for this [6]. More recently Guerrin [7] proposed a
dedicated framework to represent action with a stance close to situated action
paradigm. Hence we have decided to go for a virtual case study implemented in an
ABM and empirically grounded through previous interviews and ethnographic
analyses. The specific requirements for this virtual case include being spatially
explicit enough to generate information about the water system according to the
diversity of observations: place and type of data collected.
In this paper we present the simulation model that has been designed and
implemented and the bottlenecks we had in building it in order to be able to represent
knowledge coming from the ethnographic work and how we solved them. A first
section comes back to the pragmatic situation of drought management facing
controversies due to the diversity of indicators of water level and places to assess them
among participants to drought committees (stakeholders and county administration).
Second section defines the requirements it implies for modeling this process, with a
focus on the perception interface between the natural system and stakeholders. Third
section describes the model itself. In a last part we present first simulation outcomes in
two situations: without uses in order to validate the environmental dynamics, and with
irrigation uses and discrepancies in place of observation.
2
Drought Management Act in Practice
2.1
Main Features of Drought Management in France
Drought management at county level in France is a downscaling of a national frame
set by the 1992 water act, leaving up to the county administration to enforce it in
specific decrees adapted to local situation. These decrees set local protocols to
anticipate for and handle water shortage situations. Major droughts in 2003, 2005 and
